Proceedings of the IMAC-XXVIII

February 14, 2010, Jacksonville, Florida USA

2010 Society for Experimental Mechanics Inc.

Underwater Explosion Phenomena

and Shock Physics
Frederick A. Costanzo
Naval Surface Warfare Center Carderock Division, UERD
9500 MacArthur Boulevard
West Bethesda, MD 20817-5700
(301) 227-1650

NOMENCLATURE
C0
D
e
E
FT
G/CC
I
IN
lbs, LB, #
M/SEC
msec
P
P0
P(t)
psi
R
sec
t
W
Wi

Nominal Sound Speed

Depth of Charge
Exponential Function
Energy Flux Density
Feet
Grams per Cubic Centimeter
Impulse per Unit Area
Inches
Pounds
Meters per Second
Milliseconds
Pressure
Peak Incident Pressure
Pressure as a Function of Time
Pounds per Square Inch
Range
Seconds
Time
Explosive Charge Weight
Image Charge
Nominal Mass Density
Shock Wave Decay Constant

ABSTRACT
This paper presents a brief introduction to the basic fundamentals of
underwater explosions, including discussion of the features of explosive
charge detonation, the formation and characterization of the associated
shock wave, bulk cavitation effects, gas bubble formation and dynamics,
surface effects and shock wave refraction characteristics. Illustrations of
each of these fundamental aspects of underwater explosion (UNDEX)
loadings are made with a set of videos from a variety of experimental
testing events. In addition, analyses of associated measured loading and
dynamic response data, as well as descriptions of supporting numerical
simulations of these events are presented. At the conclusion of this
paper, each of these UNDEX effects are tied together with a summary
discussion and illustration.

INTRODUCTION
The field of underwater explosions and shock physics is both complex and fascinating.
There are many aspects of the underwater explosion event that must be studied in order to
properly understand the development and propagation of the dynamic shock loading
through the fluid. The following sections of this paper provide an introduction to the basic
features associated with the underwater detonation of an explosive charge [1], dividing the
discussion into a series of primary features. These features include: the explosion or
detonation phase, the formation of the shock wave and its effects, the secondary loading
effect known as bulk cavitation, the effects of the expanding and contracting gas bubble,
observed surface effects, and shock wave refraction effects. Each of these features will be
described in some level of detail, and then all of these features will be summarized as they
collectively form a composite illustration of underwater explosion phenomena.
Fig. 1 below illustrates a typical underwater explosion event against a full scale ship target.
In these types of events, all of the features outlined above and described within this paper
play an important role in defining the dynamic loadings that are imparted to the ship, as
well as in the visual effects that are observed in the vicinity of the event. It is the objective
of this paper to provide a detailed enough introduction of these phenomena that a deeper
understanding of the basic physics involved will be attained.

Fig. 1 Typical Underwater Explosion Test of a Full Scale Ship Target

EXPLOSION PHASE
The underwater detonation of an explosive charge can best be described as an exothermic
chemical reaction that is self-sustaining after initiation. Forming throughout the detonation
process are gaseous reactive components that are at an extremely high temperature
(approximately 3000 degrees Celsius) and pressure (approximately 50000 atmospheres).
The entire detonation process represents a rapidly propagating reaction, with propagation
speeds in the neighborhood of 25000 feet per second.
Shown below in Fig. 2 are examples of typical explosives. Presented for each explosive
listed are its explosive name, its chemical composition, as well as its specific gravity and
detonation velocity. The most common and well-known explosive type is TNT, which is
most often used as the standard when comparing energy and impulse yields of the other
types of explosives. RDX is another common type of explosive. Most of the other
explosives listed are essentially compositions of TNT, RDX and other additives in order to
produce the desired effects. The exception to this is PBXN-103, which is composed of a
variety of different elements. Explosives such as COMP B, H-6, HBX-1 and HBX-3 all have
aluminum added in order to enhance the late-time burn and thus generate greater bubble
energy. HBX-1, due to its stability and availability, is the most common explosive used by
the Navy for shock qualification purposes and full ship shock trials. As can be seen from
the chart in Fig. 2, there is a range of densities and detonation velocities associated with
this group of explosives.

EXPLOSIVE

FORMULA

DENSITY
(G/CC)

DET. VELOCITY
(M/SEC)

TNT

C7H5N3O6

1.60

6940

RDX

C3H6N6O6

1.57

8940

COMP B

RDX/TNT/WAX
59.4/39.6/1.0

1.68

7900

H-6

RDX/TNT/AL/WAX
45.1/29.2/21.0/4.7

1.74

7440

PBXN-103

AP/AL/PNC/MTN/RESOURCINOL/TEGDN
38.73/27.19/6.92/24.36/0.36/2.44

1.89

6130

HBX-1

RDX/TNT/AL/WAX

1.72

7310

HBX-3

PBX/TNT/AL/WAX
31/29/35/5

1.82

7310

Fig. 2 Examples of Common Explosives

FORMATION OF THE SHOCK WAVE

Upon detonation of the explosive charge, a very steep-fronted shock wave develops at the source
and propagates rapidly in the surrounding fluid. A schematic diagram illustrating this shock wave
propagation is shown in Fig. 3. In this diagram, one can see the shock wave propagating
spherically away from the source of detonation, while at the center of detonation a gas bubble is
forming at a much slower rate. To put things in proper perspective, the shock wave propagation
phase is on the order of milliseconds, while the bubble expansion and contraction phase is on the
order of seconds. Thus, there is about a three order of magnitude time scale difference for these
two phenomena. Such a large difference as this in the time scales does pose challenges for
some computational methods that intend to include both phases.
The shock wave propagation can be likened to the stress wave that propagates axially along an
elastic rod that has been hit on one end with a hammer. Once the shock wave has formed and
has propagated to distance beyond about 2-3 charge radii, the propagation speed remains
constant and assumes linear acoustic behavior. Inside of 2-3 charge radii, however, the
propagation is highly nonlinear. Most UNDEX applications involve intermediate or far field
scenarios, and thus for these the acoustic propagation assumption is valid. For extremely closein or near contact scenarios, the nonlinear propagation characteristics must be considered.

Fig. 3 Shock Wave from an Underwater Explosion

In Fig. 4, the far field shock wave pressure for a conventional weapon is illustrated. As can be
seen from the curve and expression for pressure as a function of time, the shock wave pressure
varies as an exponential function. As a result, the pressure function starts out at a peak value of
P0 at time zero, and decays to 1/e or about 37% of its original value in milliseconds in time.
Here is referred to as the decay constant. For the example provided in Fig. 4, which is for the
case of a 250# HBX-1 explosive charge detonated at a distance of 50 ft from the gage point, the
resulting pressure function has a peak value of about 2500 psi and decays exponentially down to
a value of about 850 psi in 0.62 milliseconds. This exponential behavior of the free-field shock
wave is extremely convenient and lends itself to straightforward computations when evaluating
shock wave impulse and energy, as will be shown later. This exponential variation of the incident
shock wave pressure is accurate for at least about one decay constant. After that point, the
incident shock wave pressure actually begins to decay at a slower rate in the tail of the shock
wave.

Fig. 4 Shock Wave Pressure History for a Conventional Weapon

Next, the variation of shock wave pressure with range will be briefly discussed. Shown in Fig. 5 is
a plot of the shock wave peak pressure vs. shock severity, expressed in the non-dimensional
form of W1/3/R, where W is the explosive charge weight and R is the standoff from the charge to
the point of interest. From this curve, it can be observed that for a large range of standoff values,
the variation of peak shock wave pressure with shock severity is linear. However, as one moves
closer to the explosive charge source location, the peak pressure variation becomes nonlinear.
This point was made earlier when describing the linear nature of the shock wave propagation for
ranges greater than about 2-3 charge radii. Inside of this range, the acoustic approximation for
shock wave propagation no longer is valid and nonlinear behavior begins to take place.

Fig. 5 - Variation of Shock Wave Pressure with W1/3/R

The description of the shock wave pressure variation with time is summarized below in Fig. 6.
Here for TNT, the relationship of peak shock wave pressure, P0, and shock wave severity, W1/3/R,
is given as a power function. A similar power relationship is provided for the shock wave decay
constant, . These relationships, referred to as shock wave similitude equations, were developed
through a series of free-field experiments conducted by the Naval Ordnance Laboratory (NOL) in
the 1950s and 1960s. In these experiments, a variety of explosive charge sizes were detonated
in the vicinity of arrays of pressure transducers positioned at various ranges from the source.
These resulting pressure measurements were then analyzed and regression analyses were
conducted to determine the best curve fits relating peak shock wave pressure and decay constant
to shock severity. The basic form of these power relations can be seen from Fig. 6 to be a
coefficient multiplied by the non-dimensional term W1/3/R raised to a power. For the decay
constant, , the power curve fit contains an additional W1/3 term.

Fig. 6 Shock Wave Pressure Functional Form

When using these power relations, one must be aware of the limits of applicability of these curve
fits. Each set of published similitude equations has associated with it the ranges of charge sizes
and standoffs that were used in the experiments to obtain the pressure data. For values of
explosive charge weight and standoff outside of these ranges the power relations may no longer
be valid. Thus, it is extremely important for the user of these power relationships to fully
understand the appropriate ranges of applicability.
Shown in Fig. 7 are the expressions for shock wave energy and impulse. For the shock wave
energy expression, the energy per unit area, also know as the energy flux density, is expressed
as 1/( 0C0) times the integral of the square of the pressure with time. For the shock wave
impulse, the impulse per unit area is expressed as simply the integral of the pressure function
with time. Due to the exponential nature of the shock wave pressure function, these expressions
for shock wave energy and impulse can easily be evaluated. Also, as was developed for the
peak shock wave pressure and decay constant, power relationships for the both the shock wave
energy and shock wave impulse were developed and are presented in Fig. 7. These empirical
expressions were developed for specific values of the upper limit of integration of usually 3, 5 or 7
times the shock wave decay constant. As a result, the analyst is cautioned when using these to
properly identify the appropriate version of the power fit for the particular problem at hand.
Finally, the concept of surface cutoff will now be described. From the diagram in Fig. 8, the direct
path of the shock wave from the explosive source to the target is illustrated with the vector
labeled R. This is the shortest path to the target and thus at the point of impingement, assuming
time is dated from first arrival of the shock wave to the target and ignoring for the moment any
shock wave reflections, the incident pressure would appear as shown in the lower of the two
pressure-time curves given in the figure. However, due to the presence of the free surface and
the fact that the shock wave propagates spherically away from the source, there is a second path
that the incident shock wave takes that intersects the water surface. Due to the significant

Fig. 7 Shock Wave Energy and Impulse

Surface Cutoff

P
t

P
Shock Wave
t

Fig. 8 Illustration of Surface Cutoff

impedance mismatch between water and air, this reflection of the compressive shock wave with
the free surface results in a tensile wave that is reflected back into the fluid. This reflected tensile
wave subsequently propagates towards the target and arrives at the point of impingement at
some finite time after the arrival of the incident shock wave. Upon arrival at the target, this tensile
wave has the effect of suddenly reducing the amplitude of the loading pressure at the target, as
shown in the upper pressure-time history shown in Fig. 8. This modification of the incident shock
wave pressure is known as surface cutoff. The time delay associated with this surface cutoff can
be computed by simply considering the respective paths of the direct and surface reflected shock
waves, and is referred to as the surface cutoff time.
As can be seen from the surface cutoff illustration in Fig.8, reflections of the incident shock wave
are extremely important and must all be considered as these reflections can significantly modify
the dynamic loading that is imparted to the target. A more detailed illustration of shock wave
reflections is presented in Fig. 9, where the effects of both a free surface and a reflecting ocean
bottom can be observed. In this example, the pressure at a selected point in the fluid, P, which
could represent a point on the target, is illustrated in the time history curve shown. Initially, the
incident shock wave compressive pressure arrives at the point of interest and results in a sharp
rise in pressure from zero gage pressure to the peak shock wave pressure. This pressure then
begins to decay exponentially in time, as shown. At the same time this is occurring, another path
that the incident shock wave travels is to the surface, and is then reflected back into the fluid as a
tensile wave. This tensile wave arrives at the point of interest delayed in time from the direct
shock wave and abruptly reduces the decayed incident shock wave pressure to some negative
value. In addition to these two loading effects, there is a third path that the shock wave can travel
and that is to the ocean bottom. Depending on the nature of the bottom material, the resulting
reflected wave can vary between a strong compressive reflection to a weak tensile reflection.
Normally, a reinforcing compressive wave is reflected and when this arrives at the point of
interest, an enhancement of the current pressure amplitude will occur, as shown in the time
history illustration in Fig, 8. This example illustrates the complex nature of identifying the dynamic
loadings associated with an UNDEX event and the importance of considering all potential
reflective sources present in the particular problem of interest.

Fig. 9 Shock Wave Reflections

BULK CAVITATION PHENOMENA

The next concept to be discussed is that of bulk cavitation [2]. Bulk cavitation occurs when a
compressive shock wave travels to the surface and is reflected back into the fluid as a tensile
wave. Since water cannot normally sustain a large amount of tension, it cavitates and thus is
transformed from a continuous, homogeneous liquid into a non-homogeneous, vaporous region.
This cavitated zone is thus incapable of further transmitting any shock disturbances in its current
state. An illustration of the region of fluid that is ultimately affected by bulk cavitation is provided
in Fig. 10. The region shown is not a snapshot in time, but an envelope showing the maximum
extent of cavitated fluid over all time. The reason that bulk cavitation is of interest is that when
the cavitated zone actually closes, due to the effects of gravity and atmospheric pressure from
above and the flow from the expanding gas bubble from below, two significant fluid masses
collide creating a water hammer effect and producing a secondary shock in the form of a
compressive pulse. This compressive pulse, known as the bulk cavitation closure pulse,
depending on the circumstances can represent a significant reloading of the target. A somewhat
exaggerated depiction of the bulk cavitation closure pulse is shown with respect to the incident
pressure and first bubble pulse. In most cases, the compressive bulk cavitation closure pulse will
be detectable but is usually of reduced magnitude with respect to the primary shock and bubble
pulses. However, certain explosive geometry scenarios can exist that bring about a more
significant and possibly damaging bulk cavitation closure pressure pulse. As a result, all UNDEX
test scenarios should explore the bulk cavitation dynamics in order to identify the possibility for a
significant target reloading.

Fig. 10 Effects of Bulk Cavitation of the Surface Layer

Illustrated in Fig. 11 are the propagation of the direct incident shock wave and its surface
reflected counterpart. As the spherical compressive shock front moves through the fluid, it loads
any structures that are within its path. Trailing behind this shock front is the spherical surface
reflected tensile wave that sweeps through and tries to reduce the total pressure in the fluid to
below the vapor pressure. As a result, a state of cavitation is produced in the fluid as this relief
wave sweeps through. The maximum extent of the bulk cavitation region that occurs over time is
depicted by the boundaries in Fig. 11 that separate the white cavitated region from the blue
uncavitated fluid. As mentioned earlier, this is not a snapshot in time.

TARGET
SHIP

SURFACE

SURFACE REFLECTION

EXPLOSION

SHOCK FRONT

Fig. 11 Illustration of the Formation of the Bulk Cavitation Region

The illustration in Fig. 12, however, with the arrows indicating the direction that the boundaries
are moving, does represent a snaphot in time of the opening and closure of the bulk cavitation
region. Here it is convenient to depict the surface reflection as emanating from an image charge
located at a distance, D, above the water surface. As the shock front and trailing surface
reflection front are passing through the region towards the right hand side of this sketch, the
forces of atmospheric pressure and gravity, along with the flow from the expanding gas bubble,
are forcing the two surfaces to collide. This point of collision, referred to as the point of first
closure, normally occurs at a point that is about of the total extent of the cavitated region. This
closure impact then propagates like a zipper going in the directions away from and towards the
charge location. The result of these cavitation closure dynamics is to produce compressive shock
impulses that reload the target. The process continues until the cavitation zone is completely
closed. As the cavitation closure pulses propagate to the surface and are subsequently reflected
back into the fluid as tensile waves, additional cavitations can occur until all of the available
energy is dissipated.

Wi
D

W
Fig. 12 Closure of the Bulk Cavitation Region

Finally, a summary of the UNDEX shock environment associated with the incident shock wave,
surface reflected wave, bottom reflected wave, and the opening and closure of the bulk cavitation
region is illustrated by the collection of snapshots in time and associated pressure-time history
shown in Fig. 13. The snapshots shown are animations from a free-field analysis of a shock
scenario using a hydrocode. For this application, the explosive charge is placed approximately
midway between the surface and a hard reflecting ocean bottom. Shortly after the explosive
charge is detonated, the snapshot shown is the upper left was generated. This schematic shows
the incident pressure wave emanating directly from the charge at the left. A black dot at the edge
of the compressive shock front represents a pressure reference point. Also seen from this
schematic is the compressive wave reflection off of the sea bottom and its propagation
throughout the fluid and the bottom material. In the upper left hand portion of this same snapshot
one can also see the formation of the bulk cavitation region (white area) as the surface reflected
wave propagates through the fluid from upper left to lower right. By analyzing the pressure at the
point of interest (black dot) associated with the snapshots at the top and bottom left of the figure,
one can see in chronological order the arrival of incident shock wave pressure and its subsequent
exponential decay, the arrival of the bottom reflected compressive wave that occurs an instant
before the arrival of the surface reflected wave, and the effect of the surface reflected wave that
reduces the pressure at this point down to the cavitation pressure.
During the next phase of response, a state of cavitation exists at the gage point as is manifested
by the constant slightly negative pressure in the corresponding time history. This is associated
with the snapshot shown in the upper right hand side of Fig. 13. This state of cavitation exists
until the atmospheric pressure, gravity and gas bubble flow work to close this region.

Finally, as the cavitated region closes, as depicted in the snapshot in the lower right hand side of
Fig. 13, the associated pressure-time history indicates the effects of the compressive closure
pulse. Although the closure pulse in this example has a low amplitude when compared with the
incident shock wave peak pressure, it does have a rather significant impulse associated with it.
This is the kind of later time impulse that can significantly reload a structure and once again
points to the importance of considering all of the aspect of the UNDEX event.

DIRECT SHOCK WAVE

CAVITATION

CAVITATION CLOSURE

BOTTOM REFLECTION

Fig. 13 UNDEX Shock Environment

GAS BUBBLE DYNAMICS

In the earlier discussion describing the detonation of an explosive charge, two primary physical
aspects of the UNDEX event were shown to develop. The first aspect of the UNDEX event, the
formation and propagation of the shock wave, was described in detail, and was observed to last
on the order of milliseconds. The second primary aspect of the UNDEX event, the expansion,
contraction and migration of the gas bubble [3], will now be described. This aspect of response
occurs within a time frame that is on the order of seconds. To introduce this aspect of response,
the diagram shown in Fig. 14 will now be explained in detail. This diagram illustrates different
phases of bubble growth, contraction and migration, and associates them with corresponding
phases of a pressure-time history.

Fig. 14 Explosion Bubble Phases and Pressure-Time History

The initial part of the time history show in Fig. 14 shows the incident shock wave pressure and its
exponential decay phase. As mentioned earlier, after the shock front propagates rapidly
throughout the medium, a gas bubble begins to expand radially outward due to the high
temperature and pressure of the explosive byproducts at its center. The gas bubble continues to
expand radially outward as the pressure inside the bubble is greater than the pressure outside of
the bubble. At some point in time, the bubble grows to the point to where the pressures inside
and outside the bubble are the same, but due to its significant outward momentum, the bubble
continues to expand radially outward. Eventually the momentum of the bubble expansion is
overcome by the imbalance between the pressure outside of the bubble and that inside of the
bubble. At this instant the bubble has reached its first bubble maximum, and there is very low
pressure inside the bubble. This is manifested in the associated pressure-time history as the long
duration negative pressure phase that exists for most of the duration of the bubble oscillation.
The bubble now begins its contraction phase, rapidly passing through the point of pressure
equilibrium and continuing on to recompress the bubble gasses. Bubble contraction continues
until the bubble cannot contract any more due to the compressibility of the gasses inside. Here
the inward contraction of the bubble is rapidly reversed causing the first bubble pulse, which is
evident in the corresponding pressure-time history.
In addition to the interplay of the dynamic forces associated with pressure imbalances and fluid
momentum, the forces of gravity and buoyancy also affect the gas bubble dynamics. As the gas
bubble expands and becomes larger in diameter, it becomes more buoyant. At the same time, as
it tries to move vertically upward, fluid drag forces resist its upward migration. Eventually, after

the first bubble maximum occurs and the bubble is contracting, it begins to migrate vertically
upward as the combination of bubble inward flow and buoyancy overcome the drag forces. This
is depicted in the diagram in Fig. 14 as a vertical movement of the center of the gas bubble to a
shallower depth. Next, the process continues with subsequent bubble expansions, contractions
and pulsations, and migrations until either the gas bubble vents to the water surface or, for
extremely deep detonations, all of the gas bubble energy is expended. For each of these
subsequent oscillations, the maximum bubble diameters are becoming progressively smaller
while the minimum bubble diameters at pulsation are becoming progressively larger. These
subsequent phases are likewise accompanied with the corresponding negative pressure and
bubble pulse aspects in the associated pressure-time history. Each of these pressures is
diminishing in amplitude with each successive pulsation.
The primary events associated with the bubble dynamics described above and illustrated by the
schematic and time-history plot of Fig. 14 will now be briefly summarized in the list below:

Summary of General Features of Explosive Gas Bubble Dynamics:

Gaseous products expand outward

Water has large outward velocity, bubble diameter increases
Internal gas pressure decreases water inertia outward
Outward flow eventually stops
Bubble contracts water flow inward
Compressibility of gas stops inward motion abruptly
Bubble pulse occurs
Process of general features repeats until bubble vents to surface or energy = 0

A photograph of an underwater explosion gas bubble that has reached its maximum radius is
presented in Fig. 15. From this view it can be observed that the bubble is nearly perfectly
spherical in shape and has not moved much vertically since its generation.

Fig. 15 Underwater Explosion Bubble at Maximum Radius

In Fig. 16, a plot of the displacement of the gas sphere as a function of time is shown for a 300-lb
TNT charge fired 50 ft below the surface. As can be observed from this plot, at the point in time
where the first bubble maximum is reached, the center of the bubble has moved very little in the
vertical direction. Its only during the contraction phase and first bubble pulse phase that the
migration of the center of the bubble becomes significant. The small series of photographs to the
right of the plot in Fig. 16 clearly illustrate the shapes and sizes of the bubble during the various
phases. Moving from top left to top right, and then from bottom left to bottom right one can see
the progression of bubble geometries throughout the initial expansion, contraction and pulsation
phases, as well as visualize the amount of vertical migration that has taken place. Also
interesting to note is the fact that during the contraction phase the bubble does not remain
spherical but instead assumes a more toroidal shape as the bottom of the bubble folds inward to
create a re-entrant jet during the bubble pulse. This geometric shape change coupled with the
associated flow assist the bubble in its vertical migration, as mentioned previously.

300-lb TNT Charge Fired 50 ft Below the Surface

Scaled Version of
Large Charge, Shallow Explosion

Fig. 16 Displacement of the Gas Sphere

Measured pressure histories from a free-field UNDEX test are presented in Fig. 17. From the
long-time playback of the recorded pressure-time history shown in the lower plot, one can
observe the initial shock wave pressure, followed by the first and second bubble pulses. For this
explosive scenario, these events all occur within one second. Above this plot are two expanded
plots that more clearly illustrate the shock wave and first bubble pulse. In these plots, one can
see the exponentially decaying nature of the incident shock wave, as well as the bell-shaped
nature of the first bubble pulse. For this explosive scenario, the duration of the incident shock
wave pressure pulse is on the order of a few milliseconds, whereas the duration of the first bubble
pulse is on the order of 100 milliseconds.

Expanded Shock Wave

Expanded 1st Bubble Pulse

SHOCK WAVE
2nd BUBBLE PULSE
1st BUBBLE PULSE

Fig. 17 Measured Pressure History

The final aspect of bubble dynamics that will be addressed in this paper is that of bubble
attraction. Fig. 18 presents a series of snapshots from a boundary element code simulation of an
explosive gas bubble interacting with a rigid wall. From this series of snapshots, one can observe
that the solid boundary acts as a flow obstruction and influences the motion of the gas bubble. In
the first snapshot, the boundary is seen not to have much effect during the initial expansion
phase. However, as the bubble begins to contract, flow on the left side of the bubble is impeded
due to the presence of the rigid boundary. Thus, as the bubble contracts its left side remains
virtually stationary while its right side begins moving towards the left. This significant effect
results in the figure shown in the second snapshot from the left, where the bubbles shape
becomes distorted and its center moves closer to the rigid plate. As the flow continues to rush
around the contracting bubble, the apparent attraction to the rigid plate becomes more evident
and the right side of the bubble begins to turn inward and jet towards the plate, as seen in the
third and fourth snapshots, respectively. The degree of attraction depends on the charge
standoff, maximum bubble radius, as well as on the size and curvature of the obstruction. Also
observed from experiments is the fact that for UNDEX generated gas bubbles, the depth and
orientation of the obstruction relative to gravity have a significant effect on bubble jet direction.

Fig. 18 Illustration of Bubble Attraction

SURFACE EFFECTS

The next feature of UNDEX phenomena to be discussed is that of observed surface effects. To
illustrate these effects, the surface phenomena observed for the detonation of a 250-lb HBX-1
explosive charge at a depth of 50-ft are sketched in Fig. 19 below. The first sketch shows the
spray dome, which is caused by the interaction of the compressive shock wave with the free
surface. As the shock wave is reflected back into the fluid as a tensile wave, the water particles
near the surface are launched vertically upward forming a parabolic shaped spray dome. This
column of water takes on a more conical shape and grows in vertical extent as the gas bubble
progresses through its first expansion phase and causes a radial outward flow.
Later in time after the water column has reached its maximum size, the first gas bubble pulse
occurs and results in a sharp protrusion through this column of a plume of water that is moving
both vertically upward and radially outward. This radial breakout of the first plume is shown in the
second sketch of Fig. 19.
As the gas bubble continues to pulsate, additional radial plumes, such as the one shown in the
third sketch of Fig. 19, will be observed. In the event that the bubble migration has occurred to
the point that during one of its pulsations it vents to the water surface, then the observed plume
associated with that pulsation will tend to be dark in color as a result of the explosive byproducts
now being released above the water surface.

Fig. 19 Surface Phenomena for 250-lb HBX-1 at 50-ft Depth

The surface effects associated with an underwater explosion are further illustrated with the series
of photographs from an actual shock test and corresponding animations from hydrocode
simulations of the same event, presented in Fig. 20. In this series of illustrations as one moves
from left to right, the progression of gas bubble states is correlated with the observed surface
effects. In the first pair of illustrations, the early expansion of the gas bubble is shown in the
simulation, and the corresponding photograph of the surface shows the spray dome caused by
the shock wave.

Bubble Maximum

Bubble Minimum

Radial Plume Breakout

Fig. 20 UNDEX Plume Above-Surface Effects

The second pair of illustrations shows the bubble at its maximum, corresponding to photo above it
of the surface state showing a conical-shaped water column. The water column has progressed
in vertical extent as can be observed in the third set of illustrations, where the gas bubble is now
shown to be at its first minimum.
The last two sets of illustrations show the corresponding simulation results and above water
photographs of the early and late phases, respectively, of the radial breakthrough of the first
plume. These surface states occur following the first bubble pulse, as was described earlier.
The extent and characteristics of the observed surface effects will vary with the explosive charge
size, depth of charge detonation, and proximity of reflecting boundaries such as the ocean
bottom. The illustration presented above, however, should enable the reader to make a general
connection between the observed events above the water surface and the shock wave and
bubble dynamics that are occurring below the surface.

SHOCK WAVE REFRACTION EFFECTS

The final element of UNDEX phenomena to be addressed is that of shock wave refraction. For
an UNDEX event, the influence of refraction on the shock wave propagation becomes of great
interest for scenarios that involve large standoff ranges where the fluid medium can have varying
thermal conditions. In this situation, the assumptions of linear acoustic propagation of the
incident shock wave begin to break down in that the thermal gradients bring about changes in the
propagation speed and thus have the effect of causing the shock wave to bend along its path
from the charge source to the target. As a result, the propagation of the incident shock wave will
be modified in terms of both speed and direction.
These shock wave refraction effects can be best described by studying the series of plots and
diagrams presented in Fig. 21. In this figure, a series of ray tracing plots generated by the
REFMS code [4] for five UNDEX tests involving long standoff ranges at decreasing standoff are
depicted. Alongside these plots are the corresponding measured sound velocity profiles from the
ocean environment at the test site. These sound velocity profiles are represented by the vertical
plots to the left of each ray tracing plot and show the variation of sound velocity in the medium as
a function of depth. From these curves, it is evident that the sound velocity values fluctuate
significantly for each test. Each ray tracing plot, on the other hand, shows the various paths that

Five Shot Sequence vs. Shallow Depth Target

Fig. 21 Observed Shock Wave Refraction Effects

the shock wave takes as it propagates from the charge source on the left side of the plot to the
target represented by the black dot to the right.
The plot set on the top left of Fig. 21 represents the condition for Shot 1, where the explosive
charge was the furthest from the target. In this case, from the ray tracing curves it can be
observed that all of the paths that the shock wave takes from the charge location bend in such a
wave that by the time they reach the range where the target is located, they completely fall below
the target. Thus, for this particular test, significantly lower shock wave effects were experienced
at the target than had been estimated assuming iso-velocity water.
As one progresses through the test series and examines the ray tracing plots associated with
Shots 2 and 3, where the target is located increasingly closer to the charge, the effects of shock
wave refraction are still evident, but these are diminishing as the standoff distance decreases.
For Shot 4, however, one can notice the fact that the bending shock wave paths actually
converge at the position of the target, focusing the shock wave effects at this location. In the ray
tracing plot for Shot 5, however, the focusing effect is less pronounced as it was for Shot 4. The
characteristics of the sound velocity profiles for these two tests were also different. The overall
effect of these variations was that even though the charge standoff for Shot 4 was larger than that
of Shot 5, the shock wave focusing effect brought about by refraction caused the levels of shock
wave energy experienced by the target to be nearly the same for these two tests.
Thus, from the example presented above in Fig. 21, refraction effects can significantly alter the
propagation of the shock wave and the energy that eventually is imparted to the target. This is
especially significant for test scenarios involving large standoff ranges and varying sound velocity
profiles. These effects must be accounted for, especially if one desires to apply scaling
techniques to the resulting dynamic loading or target response data.

SUMMARY
This report introduced and briefly described the basic elements of underwater explosions and
shock physics. From the details and examples that are presented, it is clear that UNDEX
phenomena are indeed both fascinating and complex. The illustration presented in Fig. 22 below
summarizes each of these basic features in a single diagram. From this figure one can clearly
see the various paths of the direct shock waves, the surface reflected wave, the bottom reflected
waves traveling directly through the fluid and traveling partially through the bottom material, the
bulk cavitation region, the gas bubble migration and pulsation, and the various associated surface
phenomena. All of these phenomena, along with the effects of shock wave refraction, are
important to accurately defining the dynamic loading environment that occurs throughout the fluid
medium and can affect any structure that is present.

Surface Reflected
Wave

Plume

SZ

Air

Spray Dome
Water Surface

Water
Bulk Cavitation Region
Bubble at first
Minimum
First Bubble
Pulse
Burst

Direct Shock
Waves

Bottom Reflected Wave

(Traveling entirely
through water)
Bottom Reflected Wave
(Traveling partially through
bottom material)

Ocean Bottom
Fig. 22 Summary of Underwater Explosion Phenomena

Finally, the information presented within this report provides a basic introduction to the basic
phenomena associated with underwater explosions. The objective was to provide a
detailed enough introduction of these phenomena that the reader would obtain both a
deeper understanding of the basic physics involved and a stimulated interest in the subject.

REFERENCES

1. Cole, Robert H., Underwater Explosions, Princeton University Press, 1 st Edition,

1948.
2. Costanzo, Frederick A., and John D. Gordon, A Solution to the Axisymmetric Bulk
Cavitation Problem, 53 rd Shock and Vibration Bulletin, Shock and Vibration
Information Center, Naval Research Laboratory, Washington, D.C., May 1983.
3. Snay, H. G., Hydrodynamics of Underwater Explosions, Naval Hydrodynamics
Publication 515, National Academy of Sciences National Research Council, 1957.
4. Britt, J. R., R. J. Eubanks and M. G. Lumsden, Underwater Shock Wave Reflection
and in Deep and Shallow Water: Volume I A Users Manual for the REFMS Code
(Version 4.0), Science Applications International Corporation, St. Joseph, LA.,
DNR-TR-91-15-VI, 1991.